3koozy/Predict_FBPLayer_Value

1import pandas as pd2import numpy as np3from IPython.display import display4from sklearn import preprocessing5from sklearn.neighbors import KNeighborsRegressor6from sklearn.neural_network import MLPRegressor7from sklearn.preprocessing import OneHotEncoder8from pickle import dump, load9import gradio as gr10 11# load the model12mlp_model = load(open('mlp_classifier.pkl', 'rb'))13# load the scaler14my_scaler = load(open('scaler.pkl', 'rb'))15hot_enc_scaler = load(open('hot_enc.pkl', 'rb'))16 17description = '''18This small prototype is using Big Data and AI to provide an accurate estimate of FootBall player net worth in euros based on bio info and skill level.19'''20 21def predict_value(age,height_cm,weight_kg,overall_skill,potential_skill,nationality,club):22  #pre-processing:23  numerical_features = [[age,height_cm,weight_kg,overall_skill,potential_skill]]24  catagorical_features = [[nationality,club]]25  numerical_features = my_scaler.transform(numerical_features)  26  catagorical_features = hot_enc_scaler.transform(catagorical_features).toarray()27  sample_player = np.concatenate((numerical_features[0], catagorical_features[0]), axis=0)28  #predict:29  predicted_value = mlp_model.predict(sample_player.reshape(1, -1))30  return predicted_value31 32 33demo = gr.Interface(34    fn=predict_value,35    inputs=[gr.Slider(15, 60),gr.Slider(100, 200),gr.Slider(0, 100),gr.Slider(0, 100),gr.Slider(0, 100),36            gr.inputs.Dropdown(["Argentina" , "Saudi Arabia", "England"]),37            gr.inputs.Dropdown(["FC Barcelona" , "Juventus", "Liverpool", "Al Hilal", "Al Nassr"])],38            outputs=[gr.Number(label='Net Worth (Euros)')],39            title= "TalentAI - Estimate FB Player Value (Eur)",40            description = description,41            article = "Abdulaziz Alakooz developed this prototype as part of Thkaa AI in sports contest - August 2022.")42demo.launch()